The Timeline · Back to this year on the spine
Work & income
The years over which participation in paid work falls away
A common pattern — common in the stated population; common is not required
ResearchedWhen this commonly happens
commonly 65–74 (seen 55–75)
common in the stated population; common is not required
What actually changes
- The share of people counted as working or looking for work falls steeply across these bands, and keeps falling in the oldest ones.
- Leaving paid work is spread across many years rather than concentrated on a single birthday.
- Some people are still counted in the labor force in the oldest bands the agency reports, and the figure for the very oldest band is a projection rather than an observation.
Timing
What follows describes what tends to go with each timing in a population. None of it is a recommendation, and none of it is about any particular person's life.
Earlier than the common window
What if I am early?
Leaving paid work before the common band lengthens the span funded from accumulated savings and shortens the span in which contributions and covered earnings accrue. Employer health coverage ordinarily ends with the job while public coverage keyed to age has not yet opened, so the intervening years have to be bridged from somewhere else. Withdrawals taken during a market downturn draw from a smaller base, which is the mechanism usually called sequence risk.
What tends to be harder
- A gap between employer health coverage ending and age-keyed public coverage beginning
- A longer drawdown span funded by a balance that stopped growing earlier
- Withdrawals before the age at which the additional early-distribution tax stops applying may carry that tax
Routes from here
- Marketplace coverage with income-based premium assistance, COBRA continuation of the former plan, retiree coverage where an employer offers it, or a spouse's plan
- Phased retirement, seasonal work, consulting and bridge jobs, which spread the transition rather than ending it at once
- Statutory exceptions to the additional early-distribution tax, including substantially equal periodic payments and separation from service in or after a qualifying year
Evidence: Speculative
Inside the common window
What if I am on the common path?
Leaving inside the common band is the case the surrounding system is built around: age-keyed health coverage, the claiming span for retirement benefits, and most employer plan rules all open somewhere inside or near it. Because they open at different ages rather than together, the transition is still a sequence of separate steps rather than a single event.
Evidence: Speculative
Later than the common window
What if I am late?
Continuing in paid work past the common band extends contributions, extends employer health coverage where the plan continues, and shortens the span that savings have to fund. Covered earnings in later years can replace lower years in the record from which retirement benefits are computed, and where an employer plan is still active some plan rules delay required withdrawals from that plan while the person is still working. Where benefits are already being received, amounts withheld under the earnings test before full retirement age are not forfeited: the payment is recomputed upward once that age is reached.
What tends to be harder
- Age discrimination protections apply, but roles and hours can narrow in practice well before any formal exit
- Working while receiving retirement benefits before full retirement age triggers an earnings test that withholds part of the payment
Routes from here
- Federal and state age discrimination law, enforced through the Equal Employment Opportunity Commission and state fair employment agencies
- The Social Security Administration's retirement earnings test calculator, and its procedure for reporting expected earnings so that withholding is spread across the year rather than applied in a block
- A claim filed at or after full retirement age, to which the earnings test does not apply
- Phased retirement and reduced-hours arrangements, which some employers and most public-sector plans formalize
Evidence: Speculative
Started and interrupted
What if I tried and it stopped?
Many exits from paid work are not chosen but arrive as a layoff, a restructuring, a plant or office closure, or the sudden onset of caregiving for a relative. What follows tends to be a search in a labor market where hiring at older ages is slower, and an exit that was framed as retirement only afterward.
What tends to be harder
- An unplanned end to contributions and to employer health coverage
- Longer search times, and re-employment often at lower pay or fewer hours
Routes from here
- State unemployment insurance, and the Senior Community Service Employment Program and workforce boards that fund training for older workers
- COBRA continuation, marketplace coverage with income-based assistance, or a spouse's plan for the coverage gap
- Federal and state paid and unpaid family leave provisions where the interruption is caregiving rather than job loss
Evidence: Speculative
By another route
What is the nearest viable alternative?
Leaving paid work is often not a single door but a taper: reduced hours at the same employer, seasonal or contract work, self-employment, or unpaid work that fills the same time. Income then falls in steps rather than at once, which changes when health coverage lapses and when any income-tested assistance recalculates.
Evidence: Speculative
Not at all
What if I do not want this, or cannot?
Some people work for pay until the end of their working capacity, by preference, by trade, or because the accumulated balance is small. Where the work is self-employment or a family business there is often no plan rule or employer schedule that ends it, so no exit age exists to reach.
What tends to be harder
- No planned handover of the work, and income that stops abruptly rather than tapering
Routes from here
- Retirement benefits can be claimed while still working, subject to the earnings test before full retirement age
- Business succession arrangements, buy-sell agreements and Small Business Administration transition counseling
Not doing this is a path, not a failure. Nothing on this timeline is a list of things a life has to contain.
Evidence: Speculative
Where these figures come from
Sources, with the sentence we read
“In the same year, those ages 55 to 64 had a 65.9-percent participation rate”
Golden years: older Americans at work and play — U.S. Bureau of Labor Statistics, Beyond the Numbers. data 2024 · published 2025 · checked 2026-09-03
What it measured: Civilian labor force participation rate for the band just above the prime-age group, annual average.
“In 2024, 19.5 percent of people age 65 and older participated in the labor force”
Golden years: older Americans at work and play — U.S. Bureau of Labor Statistics, Beyond the Numbers. data 2024 · published 2025 · checked 2026-09-03
What it measured: Civilian labor force participation rate for the open-ended older age band, annual average.
“The participation rate for workers age 65 to 74 is projected to be 30.2 percent in 2026, compared with 17.5 percent in 1996.”
Labor force participation rate for workers age 75 and older projected to be over 10 percent by 2026 — U.S. Bureau of Labor Statistics, The Economics Daily. data 2026 · published 2019 · checked 2026-09-03
What it measured: Projected civilian labor force participation rate for the older age bands the agency reports separately; a projection, not an observation.
“For workers age 75 and older, the participation rate in 2026 is projected to be 10.8 percent, compared with 4.7 percent in 1996.”
Labor force participation rate for workers age 75 and older projected to be over 10 percent by 2026 — U.S. Bureau of Labor Statistics, The Economics Daily. data 2026 · published 2019 · checked 2026-09-03
What it measured: Projected civilian labor force participation rate for the oldest age band the agency reports separately; a projection, not an observation.
Where this frame fails
A window flattens variation by body, by family, by place and by luck. Two people at the same age inside the same window can be in situations that have almost nothing in common, and the window says nothing about which of them anything was available to. It also describes people who have already lived this stretch — it is a record, not a forecast.